论文部分内容阅读
为了分析边坡的稳定性,利用协调粒子群算法和BP网络建立了边坡稳定性CPSO-BP预测模型。BP网络能够很好地描述边坡稳定性与其影响因素之间复杂的非线性关系,将内摩擦角、边坡角、岩石重度、边坡高度、黏聚力、孔隙压力比6个主要影响因素作为网络的输入,将边坡稳定性系数作为网络的输出。为避免BP网络陷入局部最优,利用协调粒子群算法的全局优化能力确定BP网络的连接权值和阀值,使BP网络的优势得到分发挥,达到提高模型预测精度目的。实例表明CPSO-BP模型有更好地预测精度以及将其应用于边坡稳定性预测是可行的。
In order to analyze the stability of slope, CPSO-BP prediction model of slope stability is established by using the coordinated particle swarm optimization algorithm and BP neural network. BP network can well describe the complicated nonlinear relationship between slope stability and its influencing factors. Six main influencing factors, including internal friction angle, slope angle, rock weight, slope height, cohesion and pore pressure ratio, As input to the network, the slope stability factor is taken as the output of the network. In order to avoid the BP network getting into the local optimum, the global optimization ability of the coordinated particle swarm optimization is used to determine the connection weights and thresholds of the BP network so that the advantages of the BP network can be distributed and played, so as to improve the accuracy of the model prediction. The examples show that it is feasible to predict the accuracy of CPSO-BP model and apply it to slope stability prediction.